[Talk-nz] Continuing the import of LINZ data into OSM

[Mike Kittridge]

Hi again,

I just wanted to emphasize that the waterways (as with the streets) are not
only meant to be used for cartography purposes (making the landscape LOOK
accurate). It's also about the OSM entity relationships and more
specifically in this case the river routing. The suggestions in the OSM
wiki also ensures that the routing will be entered and is of high priority.
Hence the statement, "For every river a way tagged as waterway=river, must
be drawn in the direction of the river flow (i.e. from source to sea)".
Obviously rivers change their course and we will never be perfectly
accurate in both the river centerline and the river width, but there are
some practical requirements to ensure that the OSM waterways are usable by
computational hydrologists (and other analysts) like myself.

Regards,
Mike


On Mon, Sep 13, 2021 at 12:06 PM Alan Jamieson <alwynwellington at gmail.com>
wrote:

> Kyle, thank you for your considered response.
>
> For rivers, based on my limited experience, I suggest:
>
> for those that are not physically wide (say less than 2 metres) a single
> line be mapped
>
> for wide waterways imports be a two step process:
>
>     1) Map the river banks, particularly including the variable shingle
> banks.
>
>     2) Where appropriate, introduce one or more lines to indicate the
> streams (braiding) between those banks. In real life the streams are likely
> to change their exact position over time and any any mapping should not be
> expected to show the exact number and position, rather the "complexity" of
> the subject river.
>
>
> I appreciate this is slightly at variance with the various suggestions in
> "https://wiki.openstreetmap.org/wiki/Rivers"
> <https://wiki.openstreetmap.org/wiki/Rivers>.
>
> It is offered as, for many of our rivers, water often shifts it position
> from one season to another. It my view our mapping should tend towards
> worst case and out of an abundance of caution to avoid possibly
> understating the dangers that our rivers can pose.
>
> And, so the intrepid traveller, contemplating a foot crossing of the back
> country Tauherenikau or the near urban Te Awa Kairangi, has some warning of
> what lies ahead.
>
>
> I note the "OSM tag called ref: ..." proposal.  I am not certain what is a
> contributor should do post import ex LINZ when a difference is noted.
> Should they remove the "OSM tag called ref: ... " as they edit away or ... ?
>
>
>
> On Sat 11 Sep 21 13.50, Kyle Hensel wrote:
>
> Hi Alwyn,
>
>
>
> I agree that vegetation succession is a problem in OSM - I've seen bush
> mapped as natural=scrub 10 years ago which has grown so much that it should
> now be tagged as natural=wood.
>
>
>
> In terms of landuse and vegetation, only the tree-row layer remains to be
> imported. All other landuse/vegetation layers are over 95% complete.
>
> The other remaining layers are mostly human-made features (like masts,
> fences, quarries), and the big one: rivers.
>
>
>
> There is still an issue that some of this data will be out of date in a
> few years time.
>
> To mitigate this, one thing we are doing differently this time is adding
> an OSM tag called ref:linz:topo50_id to every imported feature.
>
> This tag contains the ID used by LINZ to identify that node or area.
>
>
>
> This will allow mappers in 5 or 10 years time to compare the data in OSM
> with LINZ's data, to identify which features have been removed, edited, or
> recently added.
>
> We know it’s possible to efficiently compare LINZ and OSM’s data using the
> ref:linz:topo50_id tag beacuse we already do this monthly for street
> addresses.
>
>
>
> I hope this will prevent the issue you described from occurring with these
> new layers.
>
> Let me know if you’ve got questions
>
>
>
> Kyle
>
>
>
> *From: *Alan Jamieson <alwynwellington at gmail.com>
> *Sent: *Friday, 10 September 2021 16:16
> *To: *talk-nz <talk-nz at openstreetmap.org>
> *Subject: *[Talk-nz] Continuing the import of LINZ data into OSM
>
>
>
> My limited experience is to note elements typically dated to 2012 or
> earlier
>
> For items that refer to the natural environment, such a rivers and
> vegetation, I regulalry note recent aerial imagery has ''moved on'' from
> what was imported .
>
> With rivers the import might have included gravel banks. Typically these
> change their position each wet season.
>
> In my experience, vegetation, especially in places with little or no
> human activity, openings and scrub had become, or was tending to natural
> wood, or similar.  And many areas marked as ''managed forest'' tended to
> be pine plantations that had now passed their use by date (were not in
> fact managed).
>
> My suggestion is imports be only for the built environment.
>
> --
> nga mihi
>
> Alwyn
>
>
> _______________________________________________
> Talk-nz mailing list
> Talk-nz at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-nz
>
>
>
> --
> nga mihi
>
> Alwyn
>
> _______________________________________________
> Talk-nz mailing list
> Talk-nz at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-nz
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-nz/attachments/20210913/dee273bd/attachment-0001.htm>